Data
¶
Model
¶
1. Balance data because only 10% of observations have fires
¶
2. Preprocess data (floats standardized and categories encoded)
¶
3. Estimate with XGBoost Classifier (a sequential tree classifier)
¶
Model score?
¶
Recall score: 10%
¶
Precision score: 10%
¶
Recall score: 79%
¶
Precision score: 42%
¶
Web App
¶
Web App